© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 111674
    Prev
    Next
    INDEX
    Explanations

    with specific descriptors

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron primarily detects the preposition "with," especially when it introduces features or specifications.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     to
    -1.96
    😳
    -1.63
    EdgeInsets
    -1.55
     uska
    -1.53
    🤣🤣
    -1.49
    🤦
    -1.48
    🤷
    -1.48
    vidos
    -1.46
    🙈
    -1.45
    lamabad
    -1.43
    POSITIVE LOGITS
    3
    1.87
    6
    1.76
     ponownie
    1.63
     нашей
    1.62
     một
    1.60
     estrenar
    1.55
     naszego
    1.53
     właśnie
    1.52
     our
    1.50
    9
    1.48
    Activations Density 0.063%

    No Known Activations